This paper develops nonlinear models for diesel-electric powertrains used in medium- and heavy-duty hybrid-electric vehicles and describes a new method for designing optimal controllers for highly coupled drivetrains. Current applications of heavily loaded, heavy-duty drives call for the use of comprehensive, innovative ways of handling a large variety of nonlinear phenomena in order to optimize performance and energy conversion, to shape steady-state characteristics and to meet other desired specifications. It is evident that energy conversion and optimal energy management should be integrated and thoroughly studied. This paper demonstrates the application of derived nonlinear models of drivetrain components (diesel, synchronous generator, traction motor) analyzes the nonlinear energy conversion and develops an innovative design method to control these highly nonlinear diesel-electric systems. Fundamental analysis and design methods, such as the Lyapunov stability theory and admissibility concept, are used to attain nonlinear analysis, design and motion control. The developed concept in nonlinear analysis and design allows one to synthesize optimal tracking controllers for highly nonlinear systems with minimal approximating simplifications and assumptions used in the model development. New promising analytical results are presented along with a supporting example involving a controlled, 40-ton vehicle with a hybrid-electric drivetrain. Experimental data is reported to validate the results.